The structuring of power and the composition of norms by communicative assent
mike at zelea.com
Sun Jan 4 09:59:05 EST 2009
This is the first complete draft of the theory behind Votorola
(slightly revised from last week). I'm posting it for the record.
The Web version is easier to read:
Critical comment is welcome. -- Mike
THE STRUCTURING OF POWER AND THE COMPOSITION OF NORMS
BY COMMUNICATIVE ASSENT
Revised from: Michael Allan. 2008. SourceForge.net, project
Votorola, release 0.1.12, file d/theory.xht.
1. Peer-to-Peer Voting and Communicative Assent
Introduces a medium of communicative assent for the purpose of
consensus building. The backbone of the medium is a peer-to-peer
voting mechanism that is open to continuous recasting (delegate
cascade). It differs from the conventional media of *mass* assent
in preserving the deliberative basis of consensus, regardless of
2. The Communicative Structuring of Power
Explains how the medium may function as a primary electoral
system, one in which candidates for executive office are
nominated by open, cross-party consensus. Defines the ultimate
election of a consensus candidate as an instance of communicative
action by society as a whole. Defines assent as a steering
medium, alongside money and power, and describes how it might
rationalize the relations between lifeworld and system.
Describes how the structure of assent may serve as scaffolding
for the construction of power.
3. The Communicative Composition of Norms
Explains how the medium may be combined with a peer-to-peer
medium of collaborative drafting (recombinant text), in order to
build consensus on the composition of societal norms (laws, plans
and policies). Describes how vote flow and text flow are
interwoven in the composition, such that voters and drafters are
made equal in its authorship. Suggests how a consensus norm
might be actualized by government. Provides an example from a
legislative context, in which the unofficial participation of
assembly members opens a "public bridgehead" into the
* Notes and References
1. PEER-TO-PEER VOTING AND COMMUNICATIVE ASSENT
COMMUNICATIVE ASSENT is the expression of an agreement that arises
from discussion. This section introduces a medium in which
communicative assent is formalized through voting. The voting
mechanism is a delegate cascade that is open to continuous recasting.
In a delegate cascade, a delegate is any participant who both receives
votes (like a candidate), and casts a vote of her own (like a voter).
But when a delegate casts a vote, it carries with it those received.
And so on... Passing from delegate to delegate, the votes flow
together and gather in volume - they cascade - like raindrops down the
branches of a tree.^
FIGURE 1. Cascades in tree form. The current measure of assent
for each participant is the quantity of votes received (circled
number). Vote flow is depicted by arrows, and quantified by
volume. The votes flow together until they pool at bottom, where
they are held by the leading candidates. The red numbers are the
quantities of votes held, and thus removed from circulation.
(Cascades would likely be bushier in practice, with a typical
candidate having 5-20 immediate voters.)
Every eligible participant has a single vote, and is thus a potential
voter. She may either withold her vote or cast it. She may vote for
anyone. There are no pre-declared candidates. All participants are
eligible to receive votes. All non-participants are also eligible. A
non-participant who receives a vote is thereby made a participant, in
the role of candidate.
Votes are open to recasting. If a voter changes her mind about a
candidate, she is free to withdraw her vote, or to recast it for
another candidate. Voting is intended to remain open indefinitely,
year round, with the votes shifting as new information becomes
available to the participants. The results are never final.
Votes are public. There are no secret ballots. The identities of all
voters are revealed. Anyone may trace the flow of their votes through
the cascade, and thence to the end candidates who hold them.^
Assent for each candidate is measured by the quantity of votes
received (circled numbers in figures 1 and 2). Note that a single
vote may be received by multiple delegates before it is received by
the final candidate. As each delegate or candidate receives the vote,
her measure of assent is incremented. When the vote reaches the final
candidate from whom it can flow no further, it is "held" by that
candidate. So the total of votes held by a candidate (red numbers)
has, in itself, no bearing on the measure of assent; only the total
received (circled numbers).
The typical structure of a delegate cascade is a tree, as shown in
figure 1 (above). It has a single candidate at the root, voters at
the leaves (top), and delegates among the branches in between. The
general structure however is a cyclic graph, as revealed in figure 2.
FIGURE 2. Cyclic cascade. Depicts a cascade that has formed into
a ring structure. It is nearly perfect, but a single voter from
outside has injected a vote. The vote circles until it comes to
rest with a candidate who consequently holds two votes (red
numbers). Nevertheless the assent within the ring is equal, at 6
for each candidate.
A vote never actually cycles. It flows through every candidate
exactly once, but stops before it would re-encounter a candidate for a
second time. It then remains held where it is. Consequently the
level of assent is equal for all candidates in a cycle. A vote also
stops before it encounters its original caster. Consequently a vote
for oneself has no effect.
FIGURE 3. Tight cycles. The tightest cycle is actually between
two voters (left). A cycle with a single voter (right) is a null
cycle, equivalent to a withheld vote.
Assent is an expression of agreement. Assent is formalized in the
medium by casting a vote. The vote is cast for the person who best
represents the act agreed to. The object of agreement is always an
act. For example, Juanita may propose to build a sandbox for children
in the neighbourhood park. Samantha may agree to this act, and may
formally express her agreement by voting for Juanita.
A participant has a single vote for every act that could possibly be
proposed. If Juanita also proposes that Rajiv be appointed as Park
Superintendent, then Samantha could vote for Juanita on that act, too.
The two votes - one to build the sandbox, and one to appoint the
superintendant - would have no formal connection to each other. They
would be cast in separate "elections" so to speak.
Variant acts may be proposed. Variant acts are alternatives to the
originally proposed act. When a variant act is proposed, the
participant does not gain another vote to cast. Instead she gains a
choice of which act to vote for. She may choose either the originally
proposed act, or the variant, or neither (abstaining). Thus Monika
may hear of Juanita's plan, and join Samantha in voting for it (plan
A). She might suggest, at the same time, that the sandbox ought to be
larger than Juanita is proposing (plan B). Samantha might shift her
vote to Monika as a sign of approval for plan B, or she might keep it
with Juanita. She would have a choice. The objects of the choice
(original and variant acts) are called candidates.
A candidate always has two aspects: an active aspect, and a personal
aspect. The active aspect is the proposed act. The personal aspect
is the person who proposes, or represents, or embodies the act. Thus
the building of the larger sandbox (act) and Monika (person) are one
and the same candidate. We can speak of voting for a larger sandbox,
or of voting for Monika, but the two have the same meaning. The
personal aspect of a candidate is formalized in the medium by a
personal identifier, so one always knows *who* the candidate is for
any vote that was cast. But the active aspect is only formalized
optionally, by a link. Thus Monika may have a Web page that details
the dimensions of her enlarged sandbox. She may formally link her
candidacy to that page in order that people can discover what she is
proposing, without personally asking her. But this is optional.
The separate candidates need not be mutually exclusive. We know that
delegation may enable a single vote to be received by multiple
candidates, thereby expressing agreement to multiple, simultaneous
acts. We expect those acts to be, in some sense, logically compatible
with each other. Thus by continuing to vote for Juanita, even while
proposing a somewhat different sandbox, Monika may be expressing a
hope that Juanita will eventually agree to amend her original plan.
Then, if Samantha were to shift her vote to Monika, she might be
hoping to further that amendment. So, with a single vote, Samantha
could express her assent for both candidates, and her hope of seeing
them work together.
Assent may build as new participants join in the voting. The votes
are public, so each newcomer can discover who is involved, and what
they have agreed to. The newcomer may join in the ongoing discussion
between a candidate and her voters, and may cast a vote of her own.
She may propose new candidates and invite other participants to join.
And so on. A consensus that originally formed among 2 or 3 people
might therefore grow a little larger. In principle, there might be no
limit to its size. Before going further, however, it is important to
consider how this differs from conventional voting.
FIGURE 4. Mass assent. The same two cascades as in figure 1, but
without any actual delegation. Here each voter has recast for the
consensus candidate who currently holds her vote. As a
consequence, the tree structures have assumed a star pattern,
typical of mass voting media.
Figure 4 illustrates the case in which the voters avoid delegation and
cast directly for end candidates. Instead of a tree, the result is an
imploded star pattern, typical of mass voting media. There are many
voters, no delegates, and only a few candidates. Mass voting imposes
a limit, not to the scale of assent itself, but to the scale of
discussion. Because the number of candidates is restricted, and the
number of voters is not, there is a point at which discussion between
voters and candidates is no longer possible. That point is the
boundary between communicative assent and mass assent.
As the scale of mass assent increases, newcomers will find it
increasingly difficult to discover exactly what act is being agreed
to. Beyond a small core of initial voters, still in touch with the
candidates, the others will be in the dark. Their participation will
not have enlightened them. They will not be able to ask the
candidates what is at stake, because if the candidates were to
intercommunicate with the voters they would be flooded with questions.
The only source of information will be other forms of mass media, all
of which are inherently one-way channels. With no discussion between
the voters and the candidates, the assent of the voters cannot be
characterized as communicative.^
By contrast, if assent is mediated by a delegate cascade, the newcomer
finds it easier to join. She can discover what is at stake by asking
any of the voters or delegates in the outer branches of the cascade.
She can go shopping among the delegates, using her vote as leverage to
enter discussions, and maybe to gain concessions. She can make
suggestions. She might even solicit votes herself and thus increase
her leverage, and her ability to move deeper in the cascade. It does
not matter to her if a dozen participants are involved, or a
neighbourhood, or a whole city. The discussion is no less lively for
all its extent, and no less inviting to newcomers.
The following sections will explore how such a discussion may lead to
concrete results - how the communicative assent of the participants
may be realized as communicative action by society. For example,
consider this simple scenario:
A group of young people wish to make improvements to their
neighbourhood park. They come up with a plan and begin to promote
it locally. Some of them are in disagreement and propose
alternative plans. They all share access to a medium of
communicative assent. They use it to highlight their differences
and to resolve them one by one. Eventually the whole neighbourhood
agrees on a consensus plan. The City sends a safety inspector to
the site, and trucks in some sand. With a little help, the young
people complete the improvements to the park.^
The improvements might be as simple as adding a new sandbox for
children, or as complex as renovating the athletic facilities. The
number of participants might be a hundred for a small neighbourhood,
or ten thousand for a large one. In any case, once they had reached a
rough consensus, government would be prompted to act. The reasons
*why* are hidden in the underlying details of the scenario. Before
looking at them, it will help to consider how communicative assent may
express itself in mass elections.
2. THE COMMUNICATIVE STRUCTURING OF POWER
This section explains how large scale assent in the public sphere may
be actualized, such that society as a whole becomes engaged in
communicative action. In other words, after people agree to
something, it actually gets done. This depends on the interrelations
between communicative assent and administrative power, particularly in
regard to the election of executive officers and the delegation of
Adminstrative power in modern democracies is controlled by electoral
systems. Adding a new medium to the public sphere, while it might not
change these systems, can nevertheless change who is elected through
them. Ordinarilly a candidate of an organized political party is
elected. The path to election is shown in figure 5. The candidate is
first chosen by the party members in a primary election (1), then wins
mass assent from the wider electorate (2), and finally enters office
FIGURE 5. Electoral relations between the public sphere and the
administrative system. Detailing a transformed citizenship
interchange (2a' from figure 7) as it functions in the election of
public officials. The structural difference from the original
interchange (2a) is a medium of communicative assent (green) that
is added to the public sphere.
This same election path would be open to candidates from the medium of
communicative assent. That does not make the medium, in itself, a
political party. The medium has no leader, no staff and no members.
It has no proper name, serves no particular interests, and has no
recognized status. Nevertheless it can serve many of the same
functions as a party. A voter who did not wish to support a party
candidate might use the medium to nominate her own candidate. She
could do this simply by casting a vote. Other participants might join
her, casting their own votes, and possibly nominating their own
candidates. If the medium was used in this way, then it would come to
occupy the same political "niche" as the parties, without itself
*being* a party. It would therefore be competitive with the party
system as a whole.
To be truly competitive, however, it must meet two requirements: 1)
sufficient voter turnout in the medium; and 2) faithful carriage of
votes from the medium to the general polls. First of all, its voter
turnout must be high enough to demonstrate the electoral support of
the leading candidates. It need not match the levels of general
turnout, nor even primary turnout, but it ought to be high enough that
the candidates could extrapolate the results, and thus gauge their
support among the wider electorate. Otherwise they might not bother
to register for the general election (step 1 in figure 5).
FIGURE 6. Translation of assent from a communicative to a mass
medium. Each voter first identifies the candidate who holds her
vote at the end of the cascade (red). She then re-casts for the
same candidate in the general election (bottom).
The second requirement is that the voters must faithfully carry their
votes over to the general polls on election day (step 2). They would
have to translate assent from a communicative medium to a mass medium.
Figure 6 shows what is involved in the translation with respect to a
single winner, plurality election.^ Each voter recalls the name
of the candidate who currently holds her vote in the cascade (a name
she knows well enough, or her delegate reminds her), and casts a vote
for that same candidate at the general polling station.
If these requirements are met, then it is possible for a candidate
from the medium of assent to win an election and enter office.
Furthermore, if the turnout in the medium were ever to approach or
exceed the turnout at the general polls, then the process would likely
be irreversible. Communicative elections would become an institute of
the public sphere, and mass elections would become something of a
formality. This much is assumed in the underlying details of the park
SCENARIO DETAIL 1.1.
Mae is a community leader in the neighbourhood, and the local
delegate for the Mayor. When she learns of the plans to improve
the park she takes an interest.
Mae speaks to Hal. Hal is the local delegate for the Public
Health Officer. Mae asks Hal to look into the safety issues of
the proposed plan. Hal agrees. He takes the lead in drafting a
set of safety amendments. His amendments attract the votes of
many parents in the neighbourhood. The votes are numerous enough
to ensure that safety concerns are going to feature prominently in
The young planners have a question about the delivery of the sand,
so they approach Wen. Wen is a local building contractor, and a
delegate for the Public Works Office. He explains that several
types of sand are available from the City yards. He says that
delivery, however, will depend on budgetary approval. So they add
"sand" to the budget section of their plan.
Later, when it appears that a consensus is likely to form at some
point, Mae requests approval for the plan. She does not speak
directly to City Hall, rather she speaks to her delegate - the
person she is voting for in the Mayoral election. In reply she
receives a signed email from the Comptroller of the Parks
Department, authorizing a preliminary safety inspection of the
site. Mae then forwards the authorization to Hal, who arranges
for the actual inspection. When the safety inspector arrives, Hal
guides her to the site...
In terms of critical theory, the medium of assent is rooted in the
communicative half of society known as the "lifeworld". It transforms
the citizenship interchange relation that exists between the lifeworld
and the other half of society, known as the "system" (figure 7).
Instead of receiving what Habermas characterizes as "mass loyalty"
from the lifeworld, the system receives communicative assent; instead
of responding with political decisions, it responds with action.
Decision making is thus transferred to the lifeworld, and society as a
whole is engaged in communicative action.^
CAPTION: FIGURE 7. Relations between lifeworld and system from
the perspective of the system. In the transformed citizenship
interchange relation (2a'), large scale communicative assent from
the public sphere (A'), is actualized by the administrative system
(P). Thus society as a whole is engaged in communicative action.
Modified from Habermas.^[TCA2.320]
Communicative action is action that is coordinated by discussion aimed
at mutual understanding or agreement.^ Its application to the
structure of power is the topic for the remainder of this section.
The subsequent section (3) will address the composition of norms, such
as laws, plans and policies.
With regard to power, a candidate for executive office may win
election through communicative assent (figure 5). She would then be
placed at the root of a power structure, as shown in figure 8.
FIGURE 8. Structures of assent and assented power. A candidate
is elected to executive office by communicative assent (top). She
thus becomes the root node of a power structure (bottom), which
she proceeds to assemble.
In the case of a mass election, it ordinarilly stops there. But a
communicative election may extend its effects deeper into the power
structure by influencing the selection of subordinate officers. The
rationale for the selection of subordinates will vary, but the main
qualifications are perhaps competence, compatability and patronage.
All three may be found in abundance among the principal delegates who
voted for the executive. In each case, the measure of qualification
is the number of votes carried by the delegate. Votes are the measure
of *competence* because they qualify the delegate as a second tier
winner - a runner-up in the election - and therefore qualified for the
second tier of offices. Votes are the measure of *compatability*
because their acquisition depends on the delegates working together
for the successful election of the candidate. Finally, votes are the
measure of *patronage*, because it is only by the assent of the
delegates and their voters that the candidate is brought to office;
and the rule of patronage is, "you dance with them that brought you".
FIGURE 9. Assent as scaffolding for the construction of power.
The executive (bottom) is likely to appoint her major voters to
subordinate offices (squares). They in turn are likely to appoint
their own voters. And so on.
The major voters of an executive candidate are therefore attractive,
in their own right, as candidates for subordinate offices. Any major
voter thus appointed is placed at the commanding node of a power
sub-structure, which she may further extend by her own appointments.
Again, she may look to her own voters for guidance. And so on. The
process is recursive. Assent may therefore serve as scaffolding for
the construction of power, delegate by delegate, as depicted in figure
Once a power structure is assembled, however, the roles are likely to
be reversed. Then it is the structure of power that guides, and
assent that follows. Consider the case of local leadership in
scenario 1.1. The local leader (Mae) holds neither office nor direct
power. But she does have a measure of influence as a minor delegate
in the mayoral election. She demonstrates this when she champions the
park improvement plan, and mediates its approval. The reason for her
success stems partly from the votes she carries. If her request for
official approval had been ignored or rejected off hand, she would
probably have spoken to her delegate. At stake for the delegate would
be Mae's vote and the several hundred others she is carrying from her
neighbours. If the delegate were unhelpful, then Mae could look
elsewhere in the power structure, shifting her vote until she found a
more effective channel of influence. However, she could not shift her
vote with complete freedom. Her own electoral support is based on
influence, and to retain it she must continue to direct her vote into
the power structure (figure 9, left). The general rule is that votes
of assent will be constrained to flow in channels of communication and
influence, all within the existing power structure.
To decide the issue of a *new* power structure, the constraints of the
old one must be lifted, and assent freed to resume its guiding
role.^ This would automatically occur as the end of the term
approached, and voters' thoughts inclined to the pending decision.
Their votes would then begin to shift more freely, as illustrated in
the following sequence of diagrams:
FIGURE 10. Stable assent in mid-term. Assent is generally locked
into the power structure of the incumbent executive (bottom), and
rarely shifts outside of it.
FIGURE 11. A shift of support to a co-delegate. The end of the
term is approaching, and thoughts are inclined to the pending
decision. A principal delegate shifts her vote from the incumbent
to a co-delegate.
FIGURE 12. The co-delegate declares her candidacy. She withdraws
her vote from the incumbent, and thus declares herself a candidate
for the next term of office. A voter shifts to the rival power
structure, giving it a total of 17 votes to the incumbent's 40.
Eventually, a new structure of assent will emerge, deciding the
issue of the next election, and the structure of power for the
3. THE COMMUNICATIVE COMPOSITION OF NORMS
This section explains how communicative assent may be combined with
collaborative drafting in order to build consensus on the composition
of societal norms (laws, plans and policies). It then suggests how
such a consensus might be actualized by government, with an example
from a legislative context.
Ordinarilly, we might expect the collaborative effort to be focused on
a single document to which multiple drafters push their contributions,
as shown in figure 13. Such a centralized pattern of communication is
reminescent of mass assent. Participants are constrained to a small
number of choices (one in this case), with no formal outlet to express
dissent - except perhaps by withdrawing from the process.
FIGURE 13. Typical collaborative drafting. Typical media (such
as Wikis) have a centralized pattern of communication in which
drafters push contributions to a single, central copy of the text.
On exposure to communicative assent, however, this rigid pattern is
broken apart and dissolves. Voters begin to express their agreement
with the ideas of different drafters, freely shifting their assent
from one to another. The drafters are thus encouraged to express
their ideas more clearly, and each takes to composing a separate
document that embodies her own conception of what the norm should be.
The text therefore diverges into multiple variants that co-exist side
by side, as shown in figure 14.
FIGURE 14. Broken apart and dissolved by communicative assent.
Voters express their assent (arrows) for drafters, each associated
with a variant draft of the proposed text.
Despite their differences on some issues, the drafters will often be
in agreement on others, and will collaborate by sharing useful bits of
text amongst themselves. Their lines of communication might trace a
criss-cross pattern, with no overall direction, as shown in figure 15.
Because this form of collaboration depends on a "population" of
variants, and the cross-transfer of textual snippets, it is called a
FIGURE 15. Collaboration in a recombinant text. Instead of
pushing to the center (cf. figure 13), authors transfer bits of
text from peer to peer. Their lines of communication trace a
criss-cross pattern, more-or-less at random.
However, a drafter is not only eligible to receive votes (as a
candidate), she may also cast a vote of her own, and thus share her
votes (as a delegate) with other drafters. Naturally, she will use
these votes as editorial leverage in order to disseminate her ideas as
widely as possible in the population of variant drafts; receiving her
votes will come at the price of receiving her ideas.
But the flow of votes must cascade. Consequently, a few of the drafts
will come to accumulate a preponderance of assent, here and there,
thus expressing a tentative consensus. All of the participants will
be striving to maximize their influence on the flow of text toward
these consensus drafts; while, at the same time, the consensus
drafters will be striving to attract their votes. So the contributory
text that is drawn from the "cloud" of documents will tend to flow
along lines of assent, and deposit itself in a pattern that closely
matches the consensus (and dissensus) of the participants.^
FIGURE 16. Coordinated by drafters and voters, text and assent
cascade together. Both seek consensus (bottom) where consensus
may be found. Where it may not be found, they remain apart in
dissensus (left and right).
The structure of assent is likely to fluctuate considerably during the
composition of a norm. Unlike in executive elections, it will not
become attached to a power structure. Its only attachment will be to
the evolving content of the text. Consequently, voters will be
shifting their assent here and there, attempting to influence the
course of the composition; while drafters will be shaping the text and
shunting its flow, in order to attract assent. Figures 17 and 18
describe two examples of this interplay.
FIGURE 17. Bridging consensus. A drafter copies the text of two
dissensus documents (X and Y), and crafts a third that she hopes
will bridge them (Z). Two other drafters (A and B) shift their
votes in favour of it. These defections will place pressure on X
and Y, either to open their own drafts to Z's innovations, and
solicit her vote; or to cast their own votes for Z, and thus
complete the bridge.
FIGURE 18. Shorting consensus. 1. The consensus drafter (X) has
rejected changes proposed by a contributor (T). But a
co-contributor (S) is willing to accept them. 2. T therefore
shifts her vote to the co-contributor. The co-contributor then
withdraws her own vote from X, and thus becomes the new consensus
So shifts of assent (actual and anticipated) *steer* the evolution of
the text, pushing and pulling it into shape. Two social processes
that would otherwise be separate - the building of public consensus,
and the composition of a law, plan or policy - are thus interwoven in
the communicative medium. So detailed and extensive is their
interaction, that it will be impossible to disentangle the roles of
the voters and the drafters; all will be equally authors of the norm.
And if they are numerous enough to prompt action by government
(enforcing the law, executing the plan, or following the policy), then
it will amount to communicative action by society as a whole (figure
For example, consider a legislative process. The medium of assent
might serve as an input to the legislature, feeding in bills from the
public (figure 19). It would thus function as an alternative to
FIGURE 19. Legislative relations between the public sphere and
the administrative system. Detailing a transformed citizenship
interchange (2a' from figure 7) as it would function in the
promulgation of laws. The only structural difference from the
original interchange (2a) is a medium of communicative assent
(green) that is added to the public sphere.
Such a change in legislative relations might be furthered by a
parallel change in electoral relations (figure 5). While the public
was openly voting on bills to put before the assembly, it would also
be openly voting on the seats of the assembly members. Even without
this added influence, any proposed bill that had managed to acquire a
consensus in the public would be sure to attract the attention of law
makers, and the active support of at least some of them. The easiest
way for them to express their support would be to participate, by
casting open votes of their own. Although unofficial, their votes
would nevertheless provide a nucleus of support for an eventual vote
within the assembly - and an immediate, running forecast of its likely
success. This possiblility is illustrated in figure 20:
FIGURE 20. A public bridgehead into a legislature. Members of
the public and the legislative assembly are jointly participating
in the medium of assent. Their combined votes for the bill are
currently divided (53 and 35) between two consensus drafts. A
separate tally of the in-house votes (red) tracks the likelihood
of the bill's passage should it ever be floored. At present, the
house is voting 10 and 6 (left and right), with 6 members
undecided or abstaining. If the left were to gain 2 more votes,
the bill would have a majority in the assembly.
Before the bill could be floored, the assembly would require an
official version of the text. This might be obtained by freezing the
consensus draft for a period of time. No modifications would be
permitted during this time, but the voters would remain free to shift
their votes. If the consensus nevertheless held, then the assembly
could proceed to ratify the bill. It would then become law. In this
way, law making might be opened to the public without requiring any
structural changes to government. The changes would be entirely in
the public sphere.
NOTES AND REFERENCES
[TCA1] Jürgen Habermas. 1981. The Theory of Communicative Action.
Volume 1. Reason and Rationalization of Society. Translated
by Thomas McCarthy, 1984. Beacon Hill, Boston.
[TCA2] Jürgen Habermas. 1981. The Theory of Communicative Action.
Volume 2. Lifeworld and System: a Critique of Functionalist
Reason. Translated by Thomas McCarthy, 1987. Beacon Hill,
 Recursive transfer of votes is described by Carroll (1884) for
the purpose of constituting an assembly. Rodriguez and
Steinbock (2004) describe a method for more general purposes.
In both of these methods, the votes are alienable.^, 
Methods with inalienable votes are described by Lomax (2003) for
the purpose of constituting an assembly, and by Allan (2007) for
direct legislative voting. In both of these methods, voting is
continuous and votes are therefore inalienable. Any vote may be
withdrawn or shifted at will. It therefore remains a faithful
expression of the voter's assent, despite the fact of
Another method is Brin and Page's algorithm for ranking Web
 A secret ballot would be a defence against vote buying, because
it prevents a buyer from verifying compliance. With her vote
hidden, the voter may take the money and vote as she pleases.
So vote buying would be a poor investment.
A continuously recastable vote offers a similar defence.
Although the vote is public and compliance may be verified,
there is no guarantee of *continued* compliance. The voter may
take the money from one side, then shift her vote and take it
from the other.
 To emphasize the contrast between the communicative and the
mass, consider C. W. Mills's definitions of "public opinion" and
In a *public*, as we may understand the term, (1) virtually
as many people express opinions as receive them. (2) Public
commununications are so organized that there is a chance
immediately and effectively to answer back any opinion
expressed in public. Opinion formed by such discussion
(3) readily finds an outlet in effective action, even
against - if necessary - the prevailing system of authority.
And - (4) authoritative institutions do not penetrate the
public, which is thus more or less autonomous in its
In a *mass*, (1) far fewer people express opinions than
receive them; for the community of publics becomes an
abstract collection of individuals who receive impressions
from the mass media. (2) The communications that prevail
are so organized that it is difficult or impossible for the
individual to answer back immediately or with any effect.
(3) The realization of opinion in action is controlled by
authorities who organize and control the channels of such
action. (4) The mass has no autonomy from institutions; on
the contrary, agents of authorized institutions penetrate
this mass, reducing any autonomy it may have in the
formation of opinion by discussion.
C. W. Mills. 1956. The Power Elite. New York. p. 303-304.
As quoted in Habermas.^
 The basic types of social action are:^[TCA1.85-86]
* Teleological action - an actor attempting to reach a goal
* Strategic action - teleological action in which success
depends on the decisions of others who do not share the
* Normatively regulated action - an actor fulfilling
expectations common to the group
* Dramaturgical action - an actor self-presenting to an audience
* Communicative action - two or more actors coordinating by
discussion aimed at mutual understanding or agreement
 Black analyzes a voting mechanism proposed by Lewis Carroll that
involves multiple levels of delegation [in what sense is
unclear, until I read the original, MCA]. Unlike a delegate
cascade, however, the votes are alienated from the original
casters. The delegates treat them "as if they were their own
Lewis Carroll. 1884. The Principles of Parliamentary
Representation. Harrison and Sons. London.
 Duncan Black. 1969. Lewis Carroll and the theory of games.
The American Economic Review. 59(2), p. 210.
 Rodriguez and Steinbock et al. describe a system of
'dynamically distributed democracy' that involves recursive
delegation. Unlike a delegate cascade, however, the votes are
alienated from the original casters. The intent is to improve
the efficiency of the decision making process by removing voters
from the direct discussions. The system might therefore be
classified as a steering medium that functions as a substitute
for communicative action. Its design and purpose are therefore
different from the medium of communicative assent that is
Marko Antonio Rodriguez, Daniel Joshua Steinbock. 2004. A
social network for societal-scale decision-making systems.
NAACSOS '04. Proceedings of the North American Association for
Computational Social and Organizational Science Conference.
Marko A. Rodriguez, Daniel J. Steinbock. 2006. The anatomy
of a large scale collective decision making system. Los Alamos
National Laboratory Technical Report. LA-UR-06-2139.
Marko A. Rodriguez, Daniel J. Steinbock, Jennifer H. Watkins,
Carlos Gershenson, Johan Bollen, Victor Grey, Brad deGraf.
2007. Smartocracy: social networks for collective decision
making. 40th Annual Hawaii International Conference on System
 Lomax describes a method of constituting an assembly by proxy.
The proxies are continuously reassignable, recursively
transferrable, and typically include voting rights in the
assembly. This method differs from a delegate cascade in its
restrictive purpose. It is not generally applicable to single
winner elections, or to direct voting on laws and other objects
Abd ul-Rahman Lomax. 2003. Beyond Politics, an Introduction.
 Allan sketches the concept of an "open legislature" in which
multiple variant bills are drafted in a medium of recombinant
text, while being simultaneously exposed to delegate-cascade
Michael Allan. 2007. Recombinant text. SourceForge.net,
project textbender, release 0.2.2, file d/overview.xht.
See also the latest version online:
 Sergey Brin and Lawrence Page. 1998. The anatomy of a
large-scale hypertextual Web search engine. Computer Networks
and ISDN Systems. 30, p. 107-117.
Their PageRank algorithm for ranking Web pages, is based on
links. A link to a page is considered a "vote" for that page.
The algorithm is recursive, allowing the vote to traverse the
target page's own links. PageRank differs from a delegate
cascade in that a single page may directly vote for multiple
pages. Votes consequently multiply or split apart, instead of
cascading together. Votes also diminish in strength (are
"dampened") as they traverse the pages.
 Jürgen Habermas. 1962. The Structural Transformation of the
Public Sphere. Translated by Thomas Burger, 1989. MIT Press,
 The young people in the park improvement scenario might be of
any age. Marcus Pivato has suggested that one characteristic of
recursive delegation (as here) is that positive contributions
are to be expected regardless of the age and abilities of the
contributors. Voting might thus be opened to children, or to
the mentally infirm, with a positive effect on the overall
Marcus Pivato. 2007. Pyramidal democracy.
 Another common type of election is based on party-list
proportional representation. In this type of election, the
voter would translate her vote by recalling the party of the
candidate who currently holds her vote in the cascade, and
casting a vote for that same party at the general polling
The party would calculate its candidate list from the cascade.
It would use an algorithm similar to that of the general
election, typically one based on the highest averages method.
For discussion of how it applies to delegate cascades, see
"Delegate cascade and proportional representation":
Applying it to all cascades would yield a general all-party list
(predicative of the final election results), which could then be
divided by affiliation into a set of specific party lists.
 In order to keep the contributory, feeder drafts in sync with
the downstream consensus, text would also have to flow upstream,
in reverse. Diff/merge tools might be useful here. The
upstream drafters would use them to maintain a selective
synchrony, filtering out any incompatible content. (Though they
assent to the consensus, their own draft may express their
preference more accurately.)
Text might be pushed downstream in a "cascade of Wikis", with
the downstream drafters editing the contributions (either
lazilly or eagerly).
Or text might be pulled with the aid of various tools; anything
from recombinant text, to diff/merge, to manual copy and paste.
 In the relations between lifeworld and system, communicative
assent appears to function as a "steering medium", alongside
money and power. A "steering medium" is defined by Habermas as
a substitute for communicative reason in its capacity to
coordinate people. "Communicative reason" is a type of
rationality characteristic of free and open discourse, that aims
at a mutual understanding and agreement among participants. It
contrasts with an "instrumental reason" that is characteristic
of administered organizations (but also of engineering, science,
and so forth), that aims at a causal understanding and mastery
of the world.^[TCA1], [TCA2]
In the case of communicate assent, however, the substitution is
purely formal. Behind the formal expression, behind every
accumulation of votes that betokens a consensus, there is an
actual consensus. This is guaranteed by the freedom of the
partipants to withdraw their votes or to shift them as they
please. So the formal expression of assent is a product of
communicative reason, and its steering capacity derives from
Habermas also characterizes steering media as alienable. This
is not true of formal communicative assent, as defined here.
Although votes are delegable in a cascade (much as power is
delegable in a power structure), they may nevertheless be
withdrawn or shifted by the original caster, with complete
 I had originally thought that the formation of deliberate vote
cycles (decision rings) would be useful in guiding power shifts. I
no longer think they are necessary. The idea is summarized in the
following diagrams, the first two of which set the stage.
FIGURE 21. Rigid assent in a power structure. Ordinarilly,
assent for an executive office is locked into the power
structure of that office, and cannot shift freely.
FIGURE 22. Unstable assent in a power vacuum. The
principal delegates signal their readiness for a decision by
withdrawing their votes from the incumbent. But the ensuing
process may be overly sensitive to their initial vote
shifts, making it chaotic. And it might be difficult to
initiate the process if a large minority still supports the
FIGURE 23. A decision ring. The principal delegates cast
their votes in a massive cycle, and thus catapult ahead of
the incumbent (52 votes to 5).
FIGURE 24. A delegate backs a candidate. The delegate
exits the ring and casts her vote back into it, lending her
support of one of the remaining candidates. The candidates
in the ring all share the same vote flow (52), and are
therefore equal in assent. But the decision can also be
quantified by the flow of votes *entering* the ring (red),
as distinct from the cyclic flow within it. The sources of
votes for this separate count are restricted to the ring
candidates and their voters. No spoiler (such as the
incumbent) can contribute to the decision, unless first
invited into the ring.
I no longer see the need of decision rings. I doubt the
problems mentioned in figure 22 could affect the results in a
long running, continuous election. Any delegate who wished to
declare her candidacy could simply withdraw her vote from the
incumbent. Others could eventually decide the issue by shifting
their own votes, or not. [MCA]
- an expression of agreement
- an object of assent
- one who receives a vote
- a type of social action that is coordinated by discussion aimed at
mutual understanding or agreement
- the expression of an agreement that arises from discussion
- a subject and object of assent
- one who is both a candidate and a voter, who both receives and
- a voting mechanism in which received votes are carried along with
- the expression of agreement that arises through mass voting and
other mass media, in which there are a relatively few candidates,
and little or no communication between voters and candidates
- a standard or pattern of social behaviour; especially one that may
be formalized, such as a law, plan or policy
- a formal unit of assent
- a subject of assent
- one who votes for a candidate
Copyright 2007-2009, Michael Allan. Permission is hereby granted,
free of charge, to any person obtaining a copy of this software and
associated documentation files (the "Votorola Software"), to deal in
the Votorola Software without restriction, including without
limitation the rights to use, copy, modify, merge, publish,
distribute, sublicence, and/or sell copies of the Votorola Software,
and to permit persons to whom the Votorola Software is furnished to do
so, subject to the following conditions: The preceding copyright
notice and this permission notice shall be included in all copies or
substantial portions of the Votorola Software.
More information about the Votorola